An Efficient Frequency Domain Based Attribution and Detection Network
People nowadays can easily synthesize high fidelity fake images with different types of image content due to the rapid advances of deep learning technologies. Detecting such images and attributing them to their generative models (GMs) is crucial. Existing deep learning methods attempt to identify an...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10855423/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832575564760743936 |
|---|---|
| author | Junbin Zhang Yixiao Wang Hamid Reza Tohidypour Panos Nasiopoulos |
| author_facet | Junbin Zhang Yixiao Wang Hamid Reza Tohidypour Panos Nasiopoulos |
| author_sort | Junbin Zhang |
| collection | DOAJ |
| description | People nowadays can easily synthesize high fidelity fake images with different types of image content due to the rapid advances of deep learning technologies. Detecting such images and attributing them to their generative models (GMs) is crucial. Existing deep learning methods attempt to identify and classify GM-specific artifacts but often struggle with content-independence and generalizability. In this paper, we observe that while GMs leave unique artifacts in the frequency domain, they are coupled with the image content. Based on this observation, we propose a novel deep learning-based solution that learns input-adaptive masks to highlight GMs’ artifacts and achieve high accuracy on the synthesized image attribution task. In addition, we observed that GMs’ artifacts in the frequency domain remain intact in sub-images of the original image, and they are even retained when the images are distorted. To further improve the accuracy of the proposed solution, we leverage the characteristics of GMs artifacts in sub-images and distorted images to make our network perform more effectively. Our evaluation results show that our proposed solution outperforms other state-of-the-art methods on unseen image types, showing great generalizability. |
| format | Article |
| id | doaj-art-093e7cf60d254035865f086602199914 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-093e7cf60d254035865f0866021999142025-01-31T23:05:09ZengIEEEIEEE Access2169-35362025-01-0113199091992110.1109/ACCESS.2025.353482910855423An Efficient Frequency Domain Based Attribution and Detection NetworkJunbin Zhang0https://orcid.org/0000-0002-3645-2733Yixiao Wang1https://orcid.org/0000-0003-2664-3605Hamid Reza Tohidypour2https://orcid.org/0000-0003-0469-8410Panos Nasiopoulos3https://orcid.org/0000-0002-2654-8096Department of Electrical and Computer Engineering, The University of British Columbia, Vancouver, BC, CanadaDepartment of Electrical and Computer Engineering, The University of British Columbia, Vancouver, BC, CanadaDepartment of Electrical and Computer Engineering, The University of British Columbia, Vancouver, BC, CanadaDepartment of Electrical and Computer Engineering, The University of British Columbia, Vancouver, BC, CanadaPeople nowadays can easily synthesize high fidelity fake images with different types of image content due to the rapid advances of deep learning technologies. Detecting such images and attributing them to their generative models (GMs) is crucial. Existing deep learning methods attempt to identify and classify GM-specific artifacts but often struggle with content-independence and generalizability. In this paper, we observe that while GMs leave unique artifacts in the frequency domain, they are coupled with the image content. Based on this observation, we propose a novel deep learning-based solution that learns input-adaptive masks to highlight GMs’ artifacts and achieve high accuracy on the synthesized image attribution task. In addition, we observed that GMs’ artifacts in the frequency domain remain intact in sub-images of the original image, and they are even retained when the images are distorted. To further improve the accuracy of the proposed solution, we leverage the characteristics of GMs artifacts in sub-images and distorted images to make our network perform more effectively. Our evaluation results show that our proposed solution outperforms other state-of-the-art methods on unseen image types, showing great generalizability.https://ieeexplore.ieee.org/document/10855423/Synthesized imageattributiondetectionfrequency domain |
| spellingShingle | Junbin Zhang Yixiao Wang Hamid Reza Tohidypour Panos Nasiopoulos An Efficient Frequency Domain Based Attribution and Detection Network IEEE Access Synthesized image attribution detection frequency domain |
| title | An Efficient Frequency Domain Based Attribution and Detection Network |
| title_full | An Efficient Frequency Domain Based Attribution and Detection Network |
| title_fullStr | An Efficient Frequency Domain Based Attribution and Detection Network |
| title_full_unstemmed | An Efficient Frequency Domain Based Attribution and Detection Network |
| title_short | An Efficient Frequency Domain Based Attribution and Detection Network |
| title_sort | efficient frequency domain based attribution and detection network |
| topic | Synthesized image attribution detection frequency domain |
| url | https://ieeexplore.ieee.org/document/10855423/ |
| work_keys_str_mv | AT junbinzhang anefficientfrequencydomainbasedattributionanddetectionnetwork AT yixiaowang anefficientfrequencydomainbasedattributionanddetectionnetwork AT hamidrezatohidypour anefficientfrequencydomainbasedattributionanddetectionnetwork AT panosnasiopoulos anefficientfrequencydomainbasedattributionanddetectionnetwork AT junbinzhang efficientfrequencydomainbasedattributionanddetectionnetwork AT yixiaowang efficientfrequencydomainbasedattributionanddetectionnetwork AT hamidrezatohidypour efficientfrequencydomainbasedattributionanddetectionnetwork AT panosnasiopoulos efficientfrequencydomainbasedattributionanddetectionnetwork |